EP1726260A2 - Method and apparatus for determining neurological impairment - Google Patents

Abstract

The eyes (4) of the person to be tested are exposed to a stimulating visual signal (S,S,S); the eye movement is measured and transferred into the adequate electronic signal. A reference signal (R) is determined from the motion after an evaluation. The eyes (4) have to follow a moving stimulus (S,S,S) in a second phase and a new signal (R) is determined by comparing the motion of the stimulus (S,S,S) and the motion of the eyes (4). The device comprises a stimuli generating unit (1,6,7), an eye observing unit (3), and a comparing and evaluation unit (6).

Description

The invention relates to a method for determining neurological impairments, in which an eye movement of a person to be tested is stimulated, the eye movement is determined and converted into a corresponding electronic signal, and a test signal is obtained from the eye movement after a comparison.

A method and a device of the type mentioned is from the U.S. Patent No. 4,889,422 known. It is essentially about the detection of a dyslexia, ie an impairment when reading texts. The patient is therefore given a fixed written text to read, which he scans with his eyes. From the way the patient reads the text, that is, whether he has difficulty understanding and therefore returns to the starting point of a line or the like, is based on empirical values on his illness.

Dyslexia is a relatively rare syndrome, but where there is a certain need to take a look at neurological impairments in a known manner, the issue is the driving ability of a vehicle driver. For this purpose, proposals have already been made in various documents, for example in the WO 89/07537 , All these systems have in common that they work relatively slowly and inaccurately; in particular, it is difficult to quickly obtain a useful test signal that provides information about the condition of the driver. Therefore, these systems are neither suitable for the on-going inspection of motorists by the road police, nor for use in the vehicle itself, because the test result is obtained so late that an effective prevention of accidents can hardly be achieved.

The invention is therefore based on the object, a method of the type mentioned in such a way that a quick statement about the condition of the person or about their neurological impairment can be obtained, in which case the term "neurological" in a very general sense to be understood that is, both physical impairments (such as fatigue of the muscles) and should include physical / neurological impairments.

It should only be mentioned that the conceivable neurological impairments can be of a very extensive nature. It is not just about the impairments caused by alcohol consumption, but also about those caused by medication or drug use, by illness or simply by fatigue and exhaustion. All of these impairments can lead to accidents, but individuals, such as fatigue, can not be determined by chemical tests, such as the known alcohol tests. Another area concerns the measurement of what is today often referred to as "fitness," such as exhaustion from excessive training of athletes (in the most general sense), but also exercises by recruits in the military, although it has happened again and again is that the performance of the recruits was misjudged by the supervisors and this led to damage, even the death of the overburdened person.

According to the invention, the above object is achieved by using a positionally variable stimulus to be followed by the eye and by obtaining the test signal from the comparison of the movement of the stimulus with the movement of the eye.

So instead of using a fixed (reading) stimulus, as in the prior art, a movable stimulus is used and the eye is forced to follow this movement. Depending on the physical / mental state (fatigue and / or exhaustion, for example, is both a physical and a mental / neurological process) of each subject, the follow the eye movement will be different.

It could be that the movement patterns differ at all, because the test person simply can not concentrate on the movement of the stimulus. In general, however, it will be the case that the eye can no longer follow with the required speed or acceleration, and in this case it is favorable if, in the comparison, the time interval between the two movements is determined and used for the test signal.

In order to avoid a deviation of the eye movement transversely to the direction of movement of the position-changing optical stimulus and thus distortions in the evaluation, it is advantageous if the optical stimulus is relatively small area, in particular substantially punctiform.

The invention finally also relates to an apparatus for carrying out the method set out above.

Further details of the invention will become apparent from the following description of an embodiment schematically illustrated in the drawing. Show it:

1 shows a first embodiment of an inventive device. the

Figures 2 and 3 are diagrams for explaining the difference of the procedure according to the present invention and that of the prior art, wherein a detail X of Figure 2 is also shown on a larger scale. and

Fig. 4 shows another embodiment of the invention.

In Fig. 1, the eye 4 is shown a person to be tested for driving ability. This person is prompted (or compelled in a corresponding manner), along a line of sight A on a perpendicular to this axis A screen 1, a display or the like. to look. Such a screen 1 may be arranged in a test cabin, but it could also be a transparent LCD display on the windshield of a vehicle for periodically testing the driver be provided so that in a test, the driver's attention is not too distracted from the traffic becomes. In Fig. 1, this screen 1 is slightly unscrewed for clarity of the representation shown on the axis A to the vertical plane.

On the screen 1 or the like. is a, preferably small-area, for example, of a circle, a rectangle or even a letter or word, in particular a point, formed optical stimulus S shown, from the position S shown in solid lines in a position S 'or S " It is clear that this is the case, for example a light source such as a light-emitting diode may be provided at the positions S, S ', S ", these light sources alternately lighting up.This solution is preferred as a simplified solution, the light sources being for example as light-emitting diodes on a screen In this case, the positions S, S ', S "each correspond to a light-emitting diode.

Basically, one is in any case relatively free in choosing the type and form of the stimulus and its positions. For example, it would be conceivable that the point S does not move (only) horizontally but vertically or horizontally and vertically, etc. (X and Y direction). This is just when using a screen, i. a two-dimensional display surface, easy and very variable possible. However, it is advantageous if the stimulus after each deflection, for example in the position S 'or S "returns to a" zero position ", which may be the position S or any other.

Therefore, if a movement is simulated by this change of position of the circle or point S, the eye 4 of the subject follows this movement with a more or less long delay. Of course, such a follow-up movement would also take place when a relatively large-scale image on the screen 1 moved. The danger here is only that the eye then alternately on different points of the image, ie transverse to the direction of movement of the stimulus S follows and thereby the test result could be falsified.

The movement of the stimulus S could be controlled, for example, by an astigmatic device connected to the screen 1 or to a projection device projecting onto this screen, for example (eg with only two positions S and S ', S' and S "or S and S") astable multivibrator. However, since the movement of the eye 4 observed, converted into corresponding signals and then to be evaluated, it is advantageous if this is not done by a separate device whose output signals would then be fed to an evaluation device 6, but from a single control device. 6 which is expediently formed by a computer. Well became already stated that it would be conceivable to provide only two different positions for the stimulus S. In such a case - and then with regular jumping of the stimulus S - the danger would be great that the test person detects the beat of the movement of the stimulus and automatically moves the eyes in the same rhythm. Therefore, it is not only advantageous to have more than two positions (it can also be more than three and the movement can also be fluent), but if additionally the location (S, S ', S ") and / or the duration of the representation of the Circle, point or the like at one location by a random number generator 7 connected to the computer or processor 6. In a concrete example, the stimulus moved back and forth between nine different positions.

The distance of the positions should not be chosen too small expedient, if it is assumed that, for example, a distance which corresponds to a Sehwinkelveränderung of the eye from 0.2 ° to 03 °, is hardly perceived. Therefore, it is preferred that the distances of the positions of the stimulus S, S ', S "correspond to a visual angle change of the eye 4 of at least 3 °, in particular of at least 5 °, for example 10 °.

An additional possibility arises from possibly combining the stimulus with different contrasts, for example by emitting a stimulating luminous dot with different brightness. In this case, it would be quite conceivable to measure the reaction of the pupil (pupil enlargement - pupil reduction), which can provide a clear indication, especially with persons under the influence of drugs. The measurement of the pupil dimension is known per se in the prior art, but it is easy to determine this dimension when using a camera with image evaluation software. This possibility is also independent of the movement of a luminous point or the like. applicable.

Another type of contrast change may be that the stimulating point is displayed on a changing background. For example, as a background, a dot pattern (or a another image) which begins to move before the appearance or before the movement of the stimulating image or point or at the moment of the change in the position of the stimulating image. Additionally or alternatively, with each change of position of the stimulus, a changed, so different or differently moving, background pattern can be shown, so as to disrupt automatic movement of the eye, which could distort the measurement result as possible. Of course, however, a background pattern (moving, still, or changing the position of the stimulus) may be beneficial because it forces the subject to visually differentiate.

Thus, while the eye 4 follows the movement of the stimulus, the eye movement is scanned or recorded in a manner known per se. The recording of eye movement is in the art - about the focusing of optical systems, such as a photo camera - known per se (cf., for example DE-A-42 15 523 or JP-A-6-331880 ), and these systems can also be used in the context of the present invention. However, it is preferred if a video camera 3 is used. It is particularly preferred if the video camera 6 is designed as a high-speed camera with a larger frame rate than 100 frames / s, preferably at least 250 frames / s and in particular in a range of 400 to 600 frames / s, for example 500 frames / s. In one specific embodiment, the camera 3 had a number of 320 x 256 pixels, but this number was increased to 1280 x 1024 in another attempt.

In order to be able to reliably detect the movement of the eye 4, two light sources 5 are preferably provided, which, however, preferably emit light in a non-visible region, in particular infrared, so as not to obstruct or influence the view of the eye 4 of the subject. Although the arrangement of the light sources 5 shown is preferred, the invention is not limited thereto. This light is supplied to the eye 4 via a partially transmissive mirror 2, this mirror is preferably designed as a dichroic mirror. Consequently the camera 3 detects the movements of the eye 4 on the basis of the movement of the iris (or a part thereof), the pupil, the sclera, etc. The image sequences thus obtained can then be processed in the processor 6, for example using an autocorrelation function or with another image evaluation system. Algo-rithm, processed and the calculated deflection of the eye as a function of time in a (shown here as a separate unit, but generally contained in the processor 6) memory 9 are stored. Alternatively, an intermediate storage of the images of the camera 3. In a further memory 8 (alternatively in the same memory 9 at reserved locations) the command signals are stored as a function of time, which the processor 6 (in turn controlled by the random number generator 7) via a line 10, as in each processor, there is an internal clock (not shown in detail) which is the same for reading the memories 8, 9, respectively , both memories 8, 9 can be read synchronously and simultaneously and thus the movements of the stimulus S, S ', S "on the one hand and of the eye 4 on the other hand can be compared over time.

Figures 2 and 3 in themselves illustrate curves of movement of an eye as known in the art. In this case, instead of a moving stimulus, the test subject is given a stationary one in the form of a text to be read in several lines. The eye of an unobstructed reader thus follows - corresponding to the stair curve of Fig. 2 - each word in the line from left to right and then returns in the next line to the beginning of the line, so that then the next stair curve is formed (the horizontal axis in this representation corresponds to the time axis).

In a disabled reader (dyslexia), however, the eye movement is not so ordered, but approximately in the manner as shown in FIG. 3. However, it is only after a while that a clear and significant deviation of the eye movement from the normal can be detected, and it requires the reading of several lines before An abnormal deviation can be diagnosed with certainty.

In contrast, the method according to the invention operates in a much shorter time range, as can be seen from the enlargement of the detail X in FIG. The stair curve S _a corresponds to the illumination (visibility) of a stimulus at a specific location and over a specific time given by the width of the rectangular curve S _a . As soon as the stimulus becomes visible in a predetermined position at time t ₀ , the eye follows it with a certain time delay. This movement of the eye is not jerky, but something slowed down and is represented by the curve M. The time at which the eye has fully grasped the stimulus will be about the time t ₁ . Between t ₀ and t ₁ is the time period t, which should normally be a few milliseconds. However, in individuals who are restricted in their ability to react by taking alcohol, medications (such as analgesics), drugs, or simply by overwork, the time t may be long enough for the eye to capture the stimulus only when it changes position again , So what is measured here, in short, is actually the reaction time (or the person's ability to respond to a stimulus) and whether it still exists in sufficient abundance.

Of course, for comparison of the computer 6, corresponding comparative data must have been stored (own memory or use of a memory area of one of the memories 8 or 9). Furthermore, it is clear that some individuals are already innately slower, which is why a certain tolerance must be taken into account. The tolerance range T can be determined statistically and will then occupy a fixed width within the enlarged representation of the time t, but circumstances may be given which make it desirable to set the tolerance width T at the computer (FIG. 1) by means of a corresponding device 11 (keypad , Reglerknopf or the like.) To make adjustable. In any case, it is clear that in the organization of the data shown with reference to FIG. 1, the movement data of the stimulus S, S ', S "corresponding to the curve S _a is taken from the memory 8 whereas the data of the movement of the eye 4 can be taken from the contents of the memory 9. By superimposing these data in the sense of the enlarged representation of the detail X of FIG. 2, the reaction time t can then be obtained.

At the end, the processor 6 then emits a test signal or a signal R which expediently corresponds to the reaction time, which is the easiest way of providing information as to whether the reaction time is still within the tolerance range T (FIG. 2) or already conspicuous (according to which If desired, in a subsequent investigation can eruieren causes). This test signal R can then be made visible on a display connected to the processor, but it may also be sufficient to trigger an acoustic signal.

Fig. 4 shows an alternative embodiment of the test arrangement to that of Fig. 1. In this case, the (expensive) dichroic mirror 2 of Fig. 1 is omitted. Instead, the eye 4 is observed by the camera 3 through a hole 12 or a free space from the screen (or light emitting diodes). Either corresponding holes 13, 14 are provided for the light sources 5, or they are simply arranged in front of the screen 1 (in principle they can illuminate the eye 4 from very different directions). The evaluation via the processor 6 takes place in the manner already described, whereby the individual parts 7-9 and 11 connected to the processor (or integrated in it) are not shown in FIG.

Thus, in one embodiment of the method according to the invention, a person to be tested is in front of a screen 1 (in the broadest sense, as discussed above), then in a first step a stimulating image, such as the point S, is placed in front of the eye 4.

At the same time or only with the first movement of the stimulus, for example from S to S 'or to S "or vice versa, in a second step, the camera 3 is put into operation with the eye illumination 5 (it has already been mentioned that the eye illumination 5 can be avoided) irritation or interference of the eye 4 preferably emits light in the invisible area) Stimulus is, as already mentioned, preferably over at least about 5 ° (seen from eye 4) and in a practical example was 10 ° deflection. From this second step, the camera 3 is running to record a first sequence of, for example, 250 images for 500 ms. Selecting an increased image capture rate above 100 fps is very important for measurement accuracy.

This second step is advantageously repeated several times while changing the movement of the stimulus. It has already been mentioned that it is particularly preferred if the movements of the point S, S ', S "from one deflection point (with respect to the center position S or another" zero position ") are not moved to another deflection point (S' or S "), but only returns to the" zero position ". In this way one obtains a number of movement stimuli, for example at least 5, preferably at least 10, but the number in itself is not subject to any limitations, only those arising from the necessity of carrying out the test in a reasonable time, which - as you can see - is already possible in a relatively short time. Thus, the number of Bewegungsstimuli can be up to 100 and more. All of these sequences of eye movements are recorded by the camera 3 and set in the evaluation device 6 in comparison to the stimulus movements.

Depending on the structure of the camera 3, in a third step, either each sequence is immediately transferred to the memory of the computer 6, or the camera 3 itself has a sufficiently large memory (for example, data in the order of a few GB, as a practical experiment 2.5 GB), whose contents are only passed on to the computer 6 after completion.

1. 1. Zunächst wird, wie bei bekannten Augenbewegungsmessungen auch (beispielsweise zum Fokussieren von Kameras), ein Orientierungsschritt durchgeführt, d.h. der gewonnene Bildinhalt wird analysiert, um gewisse Bildkonturen, -kanten,

Now in a fourth step, the evaluation is divided into several steps as follows:

1. 1. First of all, as in the case of known eye movement measurements, an orientation step is also carried out (for example for focusing cameras), ie the image content obtained is analyzed in order to obtain certain image contours, edges,

• 1. In einem praktischen und vorteilhaften Ausführungsbeispiel legte der Computer eine gedachte horizontale Linie durch den Mittelpunkt dieses Kreises, wobei sich am Umfang je ein Schnittpunkt dieser Linie mit dem Kreis ergab, wobei die so gewonnenen Schnittpunkte ein erstes Paar von zu verfolgenden Punkten bildeten.
• 2. Man kann die Genauigkeit der Messung gewünschtenfalls in einem Zusatzschritt noch etwas erhöhen, indem man sich mit diesen zwei Punkten nicht zufrieden gibt, sondern die horizontale Linie weiter verfolgt, um die Grenze zwischen Iris und Sclera (Limbus) zu suchen und damit zwei weitere Punkte zu gewinnen. Algorithmen zur Ermittlung dieser Übergänge bzw. Kanten eines Bildes sind an sich in grosser Zahl bekannt. Bevorzugt (und am genauesten) ist ein auf Differentiation beruhender Kanten-Detektionsalgorithmus. So erhält man dann - wenigstens in der der "Null-Lage" (siehe S in Fig. 1) entsprechenden Stellung des Auges - insgesamt vier Punkte (links und rechts am Pupillenrand sowie links und rechts am Limbus).
• 3. Die Bildauswertung nach den obigen Einzelschritten b) bzw. c) kann an Hand mehrerer Bildsequenzen durchgeführt werden, wenn eine einzige Sequenz unbestimmt sein sollte, doch wird es im allgemeinen so sein, dass man an Hand der ersten Sequenz all diese Punkte bestimmen kann. Daher kann in jedem folgenden Bild die Position der zwei bzw. vier Punkte - jeweils unter Anwendung des Kanten-Detektionsalgorithmus - der Bewegung des Auges nachgeführt. So erhält man für jede Sequenz vier Koordinaten-Reihen, welche die Position in der Horizontalen (es wurde bereits erwähnt, dass die Stimulusbewegung in der Horizontalen bevorzugt ist) der jeweils zu verfolgenden Punkte in Funktion der Zeit beschreiben (vgl. Fig. 2).
• 4. Da die Abbildungsgeometrie bekannt ist, können die im Einzelschritt d) gewonnenen Koordinaten in Winkel umgerechnet werden. Der dafür benötigte, a priori unbekannte Augendurchmesser (d.h. nicht der Durchmesser der Pupille selbst, der ja bereits im Einzelschritt a) ermittelt wurde) kann fürs erste geschätzt werden. Gewünschtenfalls kann dann diese Schätzung mit an sich bekannten Methoden anschliessend noch genauer gemacht werden.
• 5. So erhält man die Werte für die Augenbewegung in Funktion der Zeit, beispielsweise in Form von Kurven. Diese Werte bzw. Kurven werden dann in einem weiteren Einzelschritt so ausgewertet, dass (vgl. Fig. 2) die Latenz-Zeit T (d.i. die Zeit zwischen dem Beginn der Bewegung des Stimulus t₀ und dem Beginn der Auslenkung des Auges S_a) und allenfalls die Drehgeschwindigkeit des Auges ermittelt wird. Gewünschtenfalls lassen sich auch höhere Ableitungen der Augenbewegung zur Auswertung heranziehen, doch wird im allgemeinen die Zeit T ausreichen.
• 6. Nun kann die Zeit T (oder auch weitere Bewegungsdaten des Auges) mit einem Toleranzbereich verglichen werden. In jedem Falle erfolgt nach dem Einzelschritt f) eine Speicherung und/oder eine entsprechende Anzeige. Dies kann in Form eines Ausdruckes auf Papier oder einer Darstellung auf einem Display erfolgen. Vorteilhaft ist die Angabe eines "Scores", d.h. einer Unterteilung des möglichen Variationsbereiches in Zeitverschiebungen, die in einem Bereich von "sehr gut", "gut" etc. liegen, oder man gibt einen zahlenmässigen "Score" an, um jedenfalls so auch ungeschultes Personal auf einfache Weise zu informieren, wie die Fahrtüchtigkeit bzw. die Reaktionsfähigkeit der getesteten Person liegt. Ist diese Fahrtüchtigkeit bzw. Reaktionsfähigkeit ungenügend, dann wird es angezeigt sein, dies durch weitere, länger dauernde Tests auch zu belegen. Zum Vergleich sei bemerkt, dass der erfindungsgemässe Test im allgemeinen kaum länger als fünf Minuten dauern wird und doch bereits eine aussagekräftige Beurteilung über den Fahrtüchtigkeitszustand bzw. das Reaktionsvermögen einer getesteten Person zulässt.

transitions or the like. to identify and identify. Although it has already been said above that this could be done, for example, by the iris or the sclera, the determination of the pupil is preferred. It is particularly advantageous if the edge of the pupil is determined by means of a circular Hough transformation, because this yields the circle of the pupil of known radius and center point.

• 1. In a practical and advantageous embodiment, the computer put an imaginary horizontal line through the center of this circle, with an intersection of this line with the circle being made on the perimeter, the intersection points thus obtained forming a first pair of points to be tracked.
• 2. One can increase the accuracy of the measurement, if desired, in an additional step by not satisfied with these two points, but continues to track the horizontal line to find the boundary between Iris and Sclera (Limbus) and thus two more To gain points. Algorithms for determining these transitions or edges of an image are known per se in large numbers. Preferred (and most accurate) is a differentiation-based edge detection algorithm. Thus, one obtains - at least in the "zero position" (see S in Fig. 1) corresponding position of the eye - a total of four points (left and right at the pupil edge and left and right on the limbus).
• 3. The image evaluation according to the above steps b) and c) can be performed on the basis of several image sequences, if a single sequence should be indefinite, but in general it will be so that one can determine all these points on the basis of the first sequence , Therefore, in each succeeding image, the position of the two or four points, each using the edge detection algorithm, can track the movement of the eye. Thus, four coordinate series are obtained for each sequence, which describe the position in the horizontal (it has already been mentioned that the stimulus movement in the horizontal is preferred) of the respective points to be tracked as a function of time (see Fig. 2).
• 4. Since the imaging geometry is known, the coordinates obtained in step d) can be converted into angles. The required, a priori unknown eye diameter (ie not the diameter of the pupil itself, which has already been determined in a single step a)) can be estimated for the time being. If desired, this estimate can then be made more accurate with methods known per se.
• 5. Thus one obtains the values for eye movement as a function of time, for example in the form of curves. These values or curves are then evaluated in a further single step so that (see Fig. 2) the latency time T (di the time between the beginning of the movement of the stimulus t ₀ and the beginning of the deflection of the eye S _a ) and if necessary, the rotational speed of the eye is determined. If desired, higher derivatives of the eye movement can be used for the evaluation, but in general the time T will be sufficient.
• 6. Now the time T (or also further movement data of the eye) can be compared with a tolerance range. In any case, a storage and / or a corresponding display takes place after the individual step f). This can be done in the form of a printout on paper or a presentation on a display. Advantageously, the indication of a "score", ie a subdivision of the possible range of variation in time shifts, which are in a range of "very good", "good", etc., or one gives a numerical "score", at least so also untrained To inform staff in an easy way, how the ability to drive or the responsiveness of the tested person lies. Is this ability to drive or responsiveness insufficient, then it will be appropriate to prove this by further, longer lasting tests. For comparison, it should be noted that the test according to the invention will in general hardly take more than five minutes and yet already permits a meaningful assessment of the ability to drive or the reactivity of a tested person.

In the context of the invention numerous variants are conceivable. For example, although the measurement of the reaction time t is the simplest and most significant for the purpose, but the inventive method is also applicable, for example, for brain-damaged patients or in other neurological cases to be examined, in which case it depends on which 2-dimensional course the eye movement (this is a two-dimensional "screen" advantageous) takes, ie whether she is looking directly at the position-changing stimulus or doing so indirectly. In this case, the 2-dimensional course of the movement of the stimulus can be compared with that of the eye and a corresponding test signal can be obtained from the deviations. Instead of using a camera 3, for simpler arrangements it may also be sufficient to use a plurality of photosensitive elements (photodiodes, phototransistors), e.g. in an array, to use. Also, it has already been stated at the outset that the test is suitable for all persons who are subjected to heavy use, such as in high-performance sport, in the provision of power under extreme circumstances, such as ascent of mountains of extreme heights. If a NORMAL reaction time has been defined or measured for a specific person, an overreaction, as may occur in doping cases, for example, can also be determined with the aid of the method according to the invention. Here, therefore, the term "neurological impairment" is used in its most general sense.

Unrecognized text

Claims (10)

Method for determining neurological impairments, in which an eye movement of a person to be tested is stimulated by means of an optical stimulus (S, S ', S "), the eye movement is determined and converted into a corresponding electronic signal and the eye movement after a comparison a test signal (R), characterized in that it uses a positionally variable stimulus (S, S ', S ") to be followed by the eye (4), and that the test signal (R) is obtained by comparing the movement of the human eye Stimulus (S _a ) with the movement (M) of the eye (4) is obtained. A method according to claim 1, characterized in that in the comparison of the time interval (t) between the two movements (S _a , M) is determined and used for the test signal. Method according to claim 1 or 2, characterized in that the stimulating image (S, S ', S ") satisfies at least one of the following requirements: a) that it occupies at least three different positions over time, preferably over a two-dimensional representation surface, b ) that the positions of the stimulating image are each at least approximately in a horizontal, c) that the stimulating image (S, S ', S ") is relatively small area, in particular substantially punctiform, d) the distances of the positions of the stimulus (S, S ', S ") correspond to a visual angle change of the eye (4) of at least 3 °, in particular of at least 5 °, eg 10 °, e) the stimulus (S, S', S") is changing, for example f) the contrast of the stimulus (S, S ', S ") to its background, in particular to a background pattern, is different in different positions, for example by brightness change ng of the stimulus (S, S ', S "). Method according to one of the preceding claims, characterized in that the position and / or the display duration of the stimulating image (S, S ', S ") is subjected to a random distribution. Method according to one of the preceding claims, characterized in that a number of sequences of stimulus movements are performed and each sequence is stored. Device for carrying out the method according to one of the preceding claims, having a device (1, 6, 7) emitting a stimulating signal (S, S ', S "), a device (3) observing the movement of the eye, supplying a corresponding output signal Comparison and evaluation device (6), characterized in that the stimulating signal emitting means (1, 6, 7) is a movement and / or contrast signal can be supplied, that the comparison and evaluation device (6) both said output signal and the movement or contrast signal can be supplied, wherein the differences obtained form the test signal (R). Apparatus according to claim 6, characterized in that the eye movement observing means a video recording device, in particular a camera (3) with a larger frame rate than 100 frames / s, preferably at least 250 frames / s and in particular in a range of 400 to 600 Images / s, for example 500 images / s, which is preferably associated with an image evaluation software. Apparatus according to claim 6 or 7, characterized in that the stimulating signal emitting means (1, 6, 7) at least one of the following features: a) a screen (1) for displaying a moving image and / or one, for example, moving , Background pattern; b) a plurality of light sources, in particular light emitting diodes. Device according to one of claims 6 to 8, characterized in that a random generator (7) is provided for controlling the movement signal and connected to the comparison and evaluation device (6). Device according to one of claims 6 to 9, characterized in that an adjusting device (11) for the tolerance width (T) is provided and connected to the comparison and evaluation device (6).

Images